Power consumption on Core1106 or Ultra W

2025-05-15 13:57

Hi everyone,

I've been testing the Luckfox Pico Pro board to help me with the Luckfox Core1106 board,

The difference is NAND flash instead of eMMC, and no WiFi module. With a simple script that puts the Pro into sleep mode, we managed to reduce power consumption to 0.11W (22mA @ 5V).

Now, I want to achieve the same power consumption on the Luckfox Core1106 board, but so far, it seems to draw more power - 0.30W (56mA @ 5V).
The only differences I can see are:

* Pico Pro uses NAND flash, while Core1106 uses eMMC
* Pico Pro has no WiFi module, while Core1106 does

I did check by checking powers that the NAND flash does actually goes to standby where it draws around 10uA of power so the 0.11W are consumed only from RV1106. Thus we assume that 0.17W (0.28W - 0.11W) is coming from either Wifi module or the eMMC.

We've also realised that Pico Ultra W behaves in a similar way as the Core1106 and has the same components and also has a similar power consumption so:

1. We would like to know if anyone could power up their Pico Ultra board (not Pico Ultra W board) - which is without the WiFi chip and measure the power consumption it requires in the sleep mode when this is run:

Code: Select all

echo mem > /sys/power/state

2. Are there specific configurations, hardware modifications, or software adjustments that I can apply to the Core1106 to achieve the same power consumption as the Pico Pro or at least as low as possible?

Thanks!

2025-05-16 3:08

Hello,

Since I currently don’t have access to a Luckfox Pico Ultra, I used the Luckfox Pico 86Panel-0408 (without Wi-Fi module, 8GB eMMC) to conduct sleep mode testing. The measured power consumption was 0.12W.

Due to the limited accuracy of my testing tools, I used data from the Luckfox Pico Max as a reference for comparison. The Pico Max consumed 0.11W in sleep mode.

I also tested the Luckfox Pico Ultra W.The following results are for reference only, as there may be some measurement error:

Luckfox Pico 86Panel-0408: 0.12W

Luckfox Pico Ultra W: 0.24W

Luckfox Pico Max: 0.11W

The Wi-Fi module on the Luckfox Pico Ultra W is controlled via GPIO1A2. If GPIO1A2 outputs a low level, the power consumption becomes similar to that of the regular Pico Ultra. The Core1106-1408/1208 series also uses this pin.

Quick command to disable Wi-Fi module:

Code: Select all

io -4 0xFF530000 0x00040000  # Set GPIO1A2 to output low

2025-07-26 16:21

Hello,

since we weren't able to achieve a low-power requirements of our project we implemented a MOSFET transistor in front of main power supply which cuts the power once the procesor has been halted.

However - since we'll be powering the processor on and off approximately once per 5 minutes - this will exceed 100k cycles in a year. How unhealthy is that for the unit and how many cycles can we actually expect before severely damaging the module? Is there any better way to achieve the actual low-power of the module. The RV1106 was advertised as 30mW for 4M @1fps video, and in DDR sleep the power should go down to less than 4mW, but we're nowhere that with your board.

This http://ibb.co/99SYfk1Z is the photo taken from Rockchip stand.

2025-08-04 2:48

Our board was not designed with low power consumption in mind. If you have strict design requirements for low power consumption, it is recommended to redesign the hardware solution. Additionally, the minimum 4mw you provided is the data for RV1106B. I'm not sure if RV1106G2/G3 can achieve this.

2025-08-11 13:21

luckfox_prototyper wrote: ↑2025-07-26 16:21 Hello,

since we weren't able to achieve a low-power requirements of our project we implemented a MOSFET transistor in front of main power supply which cuts the power once the procesor has been halted.

However - since we'll be powering the processor on and off approximately once per 5 minutes - this will exceed 100k cycles in a year. How unhealthy is that for the unit and how many cycles can we actually expect before severely damaging the module? Is there any better way to achieve the actual low-power of the module. The RV1106 was advertised as 30mW for 4M @1fps video, and in DDR sleep the power should go down to less than 4mW, but we're nowhere that with your board.

This http://ibb.co/99SYfk1Z is the photo taken from Rockchip stand.

You and I will probably both be dead before wear becomes an issue. If you use the emmc only for the OS, mark it as read only and you're golden.

As for the 30mW for 1fps video, that's highly unrealistic (extremely so). That's 10mA @ 3V3 - the power consumption you can expect from a MCU, not an MPU. Taking a STM32, for example, clocked at 160 Mhz (recent families, like the U5) burns around 15 mA @ 3V3 in a while(1) loop. It's next to impossible to find a MPU with a lower power consumption in any kind of active mode.

This SoM (sensor included) burns ~ 900 mW 1080p30 + software AAC encoding + WiFi streaming (access point) + local storage on a SD card. Which is pretty much half of any other MPU I've ever touched, from Texas Instruments to NXP.

2026-05-15 20:30

On pico Max I get 30mA@5V (no camera connected) using my FNIRSI 48P. This is only HALF the running mode current (idle), which is weird. On most boards I have tested, suspend current is typically 1/5 the running (idle) current. As for the 30mW for 4M @1fps video, it must refer to AOV feature, like the one here https://zhuanlan.zhihu.com/p/1967544158798459292. 4mW in DDR sleep is just impossible